Biochemistry 1984, 23, 66 14-66 18
6614
Optically Detected Magnetic Resonance of Escherichia coli Glutamic Acid Specific Transfer Ribonucleic Acid and Its Anticodon-Anticodon Complex with Yeast Phenylalanine-Specific Transfer Ribonucleic Acid? Mohammad-Reza Taherian, Kuen-Fai Steven Luk, and August H. Maki*
ABSTRACT:
The low-temperature phosphorescence and the optically detected magnetic resonance (ODMR) spectra of Escherichia coli and its anticodon-anticodon complex with yeast tRNAPhcare reported. The ODMR signals are assigned to the modified base 5 - [ (methylamino)methyl]-2-thiouracil (mnmss2U) located at the "wobble" position of the anticodon. The zero-field splittings (zfs) are larger than those found for 1-methyl-2-thiouracil previously studied in a 1-methyluracil host system by ODMR. They are comparable, however, to those found for neat, polycrystalline 1-methyl-2-thiouracil and for the latter dissolved in ethyl
acetate solvent. In the polycrystalline sample, five traps with widely varying zfs are assigned. A very large (1 5-25%) reduction in the 1 0 1 parameter of mnmSs2Uis found to occur on formation of the anticodon-anticodon complex with yeast tRNAPhe. Additional ODMR signals found in the complex are assigned to the wybutine base of yeast tRNAPhe. The extreme sensitivity of the zfs parameters of s2U to environmental perturbations is ascribed to the variable involvement of the heavy atom containing chromophore, C=S, in the triplet-state wave function, which is largely localized on the C=C-C=O portion of the molecule.
x i o u r a c i l s frequently occur as modified bases in the tRNAs of Escherichia coli. The modified base s4U' is found at position 8 from the 5'-end of most tRNAs from E. coli (Singhal & Falls, 1979). The phosphorescence emission of the base s4U has been used (HClEne et al., 1968; HClEne & Yaniv, 1970) in spectroscopic studies of E. coli tRNAs at 77 K. In order to further extend the usefulness of thiouracils as probe molecules of tRNA structure, we have used optical detection of magnetic resonance, ODMR (Clarke, 1982), to investigate the phosphorescent triplet states of m1s2U,m1s4U, and m1s2s4U(Taherian & Maki, 1981a; Taherian, 1981). The zfs of these triplet states is found to be unexpectedly large, with 1 0 1 increasing in the order s2U < s4U < s2s4U. The large value of the zfs in s4U can be understood in terms of the internal heavy atom effect introduced by the sulfur atom (Taherian et al., 1982). In a previous publication (Taherian & Maki, 1981b) we reported the ODMR study of E. coli tRNAIVa'. In that work, we focused on the ODMR signals of s4U that is present at the 8-position (5' 3') of that tRNA. It was found that the zfs of the s4U triplet state were sensitive to conformational changes resulting from removal of Mg2+ from solution. The phosphorescence spectrum of tRNAIVal was found to be sensitive to the excitation wavelength. Blue-shifted emission produced ODMR signals that were assigned to the modified base cmo5U, which is located at the 5' end of the anticodon triad. In this paper we extend our ODMR investigations to another E. coli tRNA, tRNAZGlU, which is one of the few tRNAs from this organism that does not contain the base s4U. Instead, tRNAzGluhas the modified base 5-[(methylamino)methyl]2-thiouracil, mnm5s2U,at the 5' end of the anticodon, often called the "wobble" position. The presence of a modified base at this position appears to be a well-documented structural property of tRNA molecules, having only three known ex-
ceptions (Singhal & Falls, 1979). Bases in this position are modified in a characteristic manner; a uridine at this position, for instance, is often modified by sulfur substitution at the 2-position and for attachment of a bulky group at position 5 (Nishimura, 1972; McCloskey & Nishimura, 1977). Whereas we found previously that the zfs parameters of s4U were relatively insensitive to changes in local environment, the zfs of s2U are found in the present study to be extremely sensitive to the local environment. The zfs of mnmss2U in E. coli are found to differ greatly from the model system of m1s2Uin crystalline mlU studied earlier (Taherian & Maki, 1981a). The variation of the zfs is indeed a genuine environmental perturbation, and not due to the substituent in the 5-position, since we find also that the large zfs changes are mimicked by m's2U when dissolved in ethyl acetate or when studied as a neat polycrystalline sample. We also have studied the effect on the zfs of mnmSs2Uof forming the anticodonanticodon complex between yeast tRNAPhe and E. coli whose anticodons are complementary (Eisinger, 1971a,b). This complex has been studied previously by ODMR spectroscopy (Luk, 1975).
-
+ From the Department of Chemistry, University of California, Davis, California 95616. Receiued March 17, 1984. This work was supported in part by grants from the National Science Foundation and the National Institutes of Health.
0006-2960/84/0423-6614$01.50/0
Materials and Methods m1s2Uwas prepared by published procedures (Vorbrueggen & Strehlke, 1973). The product was chromatographed over
a silica gel column by using ethyl acetate as the eluent. White crystals, which separated after concentrating the solution, were purified further by recrystallization from methanol. Phosphorescence and ODMR spectroscopic measurements were made on the neat polycrystalline material and on lo4 M I Abbreviations: s"U, n-thiouracil; mls"U, I-methyl-n-thiouracil; m's2s4U, 1-methyl-2,4-dithiouracil; ODMR, optical detection of tripletstate magnetic resonance; zfs, zero-field splittings; D and E , zero field splitting parameters; cmo5U, 5-(carbomethoxy)uridine; mnm5s2U, 5[(methylamino)methyl]-2-thiouracil;EG,ethylene glycol; EPA, ethyl ether-isopentane-ethanol (5:5:2); AM-PMDR, amplitude-modulated phosphorescence-microwave double resonance; yW, wybutine; Gm, 2'methoxyguanosine.
0 1984 American Chemical Society
V O L . 2 3 , N O . 2 6 , 1984
ODMR OF TRANSFER RNAS
6615
Table I: Phosphorescence Origins and Zero-Field Splittings for Uracil and Some Sulfur Derivatives V I (GHz)' ~2 (GHz)' up (GHz)" sample origin (nm) [AVl (MHz)l [ A y z (MHz)l [AVO(MHz)l ID1 (cm-') IEl (cm-') Poly (Wb -0.65 5.48 [370] 6.13 [410] 0.194 0.01 1 r n W in mlUc 391.2 4.365 [2] 6.500 [35] 10.872 [20] 0.2895 0.0728 mk2U (neat) trap A 397.7 4.73 [21] 7.17 [200] 11.59 [130] 0.308 0.079 397.8 4.64 [21] 8.40 [130] 13.10 [160] 0.358 0.077 trap B 397.9 3.78 [230] 9.11 [130] 12.86 [140] 0.367 0.063 trap C trap D 398.0 5.66 [230] 8.22 [160] 13.84 [220] 0.369 0.094 398.2 4.69 [21] 9.87 [165] 14.52 [200] 0.406 0.078 trap E 475.0 3.001 14.41 16.57 11071 19.70 [lo01 0.605 0.0500 mls4U in mlUC 'Numbers in brackets are the line widths at half-maximum intensity. bData are from Luk (1975). The sample is dissolved in ethylene glycol-H20 and is complexed with Ag+ to enhance the phosphorescence intensity. CDataare from Taherian & Maki (1981a).
solutions in ethyl acetate (Fluka, Puris.). E . coli MRE 600 tRNA2G'uand brewer's yeast tRNAPhewere obtained as lyophilized powders from Boehringer-Mannheim. Aqueous solutions were prepared by dissolving the tRNAs in 10 mM MgCl, and annealing the solution for '/, h at 80 "C. Under these conditions, the tertiary structure will be melted (Crothers et al., 1974). The samples were cooled slowly, over a period of 30 min, to room temperature and subsequently dialyzed against cacodylate buffer (30 mM cacodylate, 100 mM NaCl, 10 mM MgC12, pH 7) for 2 days at room temperature. Other samples of the tRNAs were prepared similarly but omitting the thermal melting and annealing steps. The phosphorescence, the ODMR, and the effects of association via anticodon complexes were indistinguishable from those measured for the heated tRNAs. Although we cannot rule out completely the formation of some cleavage fragments by heating with MgZ+, they apparently are not present in sufficient quantities to affect the spectroscopic observations. The tRNA concentrations were determined by UV absorption spectroscopy using a molar extinction coefficient of 6.3 X lo5 at 260 nm (Eisinger, 1971a) and found to be 5.4 X M (tRNA,G1U)and 1.5 X lo4 M (tRNAPhe). the tRNA solutions were mixed with equal volumes of ethylene glycol (EG, Matheson Coleman and Bell, chromatoquality) for low-temperature spectroscopic studies. The anticodon-anticodon complex was formed by mixing appropriate volumes of the aqueous tRNA solutions and keeping the mixture at 0 OC overnight before addiiton of EG. Ethylene glycol may have some influence on tRNA structure. We have found in previous work, however (Hoover et al., 1974), that a n t i c o d o n d o n recognition is not affected in the presence of 50% EG. Low-temperature phosphorescence and ODMR studies were carried out by using apparatus and methods described previously (Maki & Co, 1976; Maki et al., 1978; Taherian & Maki, 198la,b). Phosphorescence spectra were measured either at 77 K or at the temperature of pumped liquid He (1.1-1.2 K). ODMR measurements were made under the latter conditions. AM-PMDR measurements (ElSayed et al., 1970; Olmstead & ElSayed, 1974) were carried out at 1.1-1.2 K by using previously described procedures (Davis & Maki, 1982). Results ( A ) I -MethyI-2-thiouraciI (Neat Polycrystalline Sample). The phosphorescence spectrum of m's2U, shown in Figure 1, is characterized by a well-resolved 0,O-band at 398.2 nm with a full width at half-maximum intensity of ca. 50 cm-I. There is an underlying broad emission, but well-resolved vibronic bands are found at 0,O-723 and at O,O-1617 cm-'. Except for the broad underlying emission, the spectrum is very similar to that observed previously for m's2U as a guest in polycrystalline m'U (Taherian & Maki, 1981a). There is, how-
P
,
~
I
~
-
" t
e n
-
5
I
t
Y
-
I
I
,,I, , ,
,
(
,
, , , ,
,
, ,
, , , , I ,
, , ,
~
6616
TAHERIAN, LUK, A N D MAKI
BIOCHEMISTRY
Table 11: Zero-Field Solittines of E . coli tRNAqG1" and Its Anticodon-Anticodon ComDlex with Yeast tRNAPbc U, (GHz)" ~2 (GHz)" U) (GHz)" sample [ A h (MH4l [Au2 (MHz)l [Au3 (MHz)l PI (cm-') tRNAZG'" site I 4.4 [300] 9.3 [600] 12.9 [700] 0.370 site I1 4.4 [300] 11.1 [800] 14.7 [800] 0.430 anticodon-anticodon complex mnm5s2U signals 4.3 [550] 7.3 [1250] 0.3 15 yW signals 0.81 [200] 1.69 [240] 2.51 [200] 0.0700
IEI (cm-') 0.073 0.073 0.072 0.0135
"Numbers in brackets are the line widths at half-maximum intensity. P h
P h
E
0
I
5
R
E
8
r e
o
9
C I
e n
8
B
C
I
I
n
n
I
t
I
9
n
n
,
s
I
I
I
t Y
Y
0 Wavelength (nm)
FIGURE 2: ODMR spectra of neat, polycrystalline mls2U at 1.1 K. The sample is excited at 312 nm and the 0,O-band is monitored with a monochromator bandwidth of 0.96 nm. The signals are the result of 2048 scans with each range swept at a rate of 12.5 MHz/ms.
assignments of zfs to individual sites by optical selection because of the broad, unstructured character of the phosphorescence. It is clear, however, that several discrete emitting sites are present having variations in zfs even larger than those observed in neat m1s2U. ( C ) E . coli tRNAZG'".The phosphorescence spectrum of E . coli tRNAzGluis independent of excitation wavelength, indicating that there is only one emitting base. The spectrum, shown in Figure 3, is broad and poorly resolved, resembling that of m1s2U in ethyl acetate. The origin of the phosphorescence is at 388 nm, somewhat to the blue of m1s2Uin ethyl acetate. The decay of the phosphorescence at 1.2 K when monitored at 406 nm with 3-nm slits could be fit to a sum of three exponential components (3.8 ms, 66%; 33 ms, 17%; 200 ms). These decay lifetimes are similar to, although somewhat shorter than, those found (Taherian & Maki, 1981a) for mlsZU in mlU by using transient ODMR methods. The phosphorescence of E . coli tRNA2G1uis accordingly assigned to the minor base, mnm5s2U,which occupies the wobble position of the anticodon, 5'-mnm5s2U-U-C-3'. The ODMR spectrum of E. coli contains a prominent relatively low frequency line at 4.40 GHz. This frequency is close to ul frequency of m1s2Ugiven in Table I. This signal is accordingly assigned as the v1 transition of mnm5s2U in the tRNA. Higher frequency transitions are found in the region 8-18 GHz; these are shown in Figure 4. Four broad signals can be resolved in this frequency range. One possible assignment of these signals is to 1D1-lEl and ID1 + IEI transitions from two resolved sites. The value of E is nearly the same for both sites, so only one 21EJ signal is resolved at ca. 4.40 GHz. Thus, the two sites are distinguished by a difference in the value of D . The ODMR frequencies and zfs parameter assignments are included in Table 11. It should be noted that this assignment of the transitions of E .
FIGURE 3: Phosphorescence spectra of (a) E . coli tRNAZG'",(b) yeast tRNAPhc,and (c) the anticodonanticodon complex of the two tRNAs. The temperature is 4.2 K, and excitation is at 310 nm.
1
I'
1
I
I II
II I I II I '
I I
I II I II ' I ' ' I
' ''I ' I
I
III
'I
I ' I '
I
4
7 8 9 1 0 1 1 1 2 l 3 1 4 1 5 1 6 1 7 l E Frigunacy (MI)
4: ODMR spectra of E. coli tRNAZG1" at 1.1 K. The sample is excited at 312 nm with the phosphorescence monitored at 410 nm. The region between 8 and 12 GHz is swept at 8.8 MHz/ms, and 3600 scans have been averaged. The 12-18-GHz region is swept at 11.2 MHz/ms and is averaged for 1260 scans. FIGURE
coli does not provide particularly good consistency between the three ODMR transitions for which u1 + u2 = v3 should hold. The signals are broad, however, and there can be a significant apparent breakdown of this relationship in heterogeneous samples (Maki & Co, 1976). In order to confirm the assignments of the two highest frequency ODMR transitions to distinct emitting sites of the tRNA, we measured the AM-PMDR spectra (ElSayed et al., 1970; Olmstead & ElSayed, 1974) obtained by coherently modulating each of these transitions in turn. The AM-PMDR spectra obtained each resemble the phosphorescence of the tRNA, but the
VOL. 2 3 , N O . 2 6 ,
ODMR OF TRANSFER R N A S P
I
1
I\
i\
I
i Y I
3
/ " " 1 " " 1 " " l " " l " " I 4
5
6
7
a
9
Frequency (GHz)
FIGURE 5 : ODMR spectra at 1.1 K of (a E. coli tRNAZG1", (b) the anticodonanticodoncomplex with tRNAbh,:tRNAZGlU = 1:1, and (c) the anticodonanticodon complex with tRNAPhe:tRNAzGIU = 2.1. The samples were excited at 310 nm, and the phosphorescence was monitored at 430 nm. The microwave sweep rate was 9.2 MHz/ms.
spectrum obtained by modulating the higher frequency ODMR signal is shifted to the red by ca. 620 cm-' relative to the other. This confirms the assignment of the two highest frequency transitions to distinct mnm5s2Usites and shows that the site with the larger ID1 has the lower triplet-state energy relative to the ground state. ( D ) Anticodon-Anticodon Complex of Yeast tRNAPheand E . coli tRNAZG'". The phosphorescence spectrum of the complex formed with the ratio tRNAPhc:tRNAzGlu = 2:l is compared in Figure 3 with those of E. coli tRNAZGluand yeast tRNAPhefor the same excitation wavelength of 310 nm. It is clear that the phosphorescence resembles that of E . coli tRNAZGluand is thus mainly due to mnm5s2U. The phosphorescence of yeast tRNAPhchas been shown previously (Hoover et al., 1974) to originate from the highly modified base wybutine (yW), which occupies the position adjacent to the 3' end of the anticodon, 5'-Gm-A-A-3'. ODMR signals of yW are found below 3 GHz (Hoover et al., 1974; Luk, 1975). The phosphorescence of the anticodon-anticodon complex formed between the tRNAs at the ratio tRNAPhc: tRNA2G1u= 1:1 contains significant long-lived components (1-2 s) in addition to the short-lived decay of mnm5s2U. The presence of these long-lived components suggests that yW contributes to the decay. This is confirmed by characteristic yW ODMR signals given by the anticodonanticodon complex, which will be described below. The effect on the ODMR signals of mnm5s2Uin the 4-8 GHz signal region when the anticodon-anticodon complex is formed is shown in Figure 5. As the complex is formed, a new ODMR signal appears at 7.3 GHz that is not present in E . coli tRNAZG1",itself. Its intensity increases as the ratio tRNAPhC:tRNAGIU is increased from 1:1 to 2: 1. At the same time, the higher frequency signals observed in tRNA2G1u (Figure 4) vanish into the noise, but the v 1 signal is unaffected. The 7.3-GHz signal is close in frequency to that assigned to the v2 transition of mlszU in polycrystalline mlU (Table I). Since this frequency is larger than those of yW or of any other bases that give ODMR signals in these tRNAs (Luk, 1975) and has a short response time characteristic of a heavy atom perturbed base, we assign it to mnm5szU as its u2 frequency in the complex. Although no v 3 transition is observed, these results suggest that the formation of the anticodon-anticodon complex leads to the reduction of 1 0 1 to a value that is comparable to that observed from m1s2U in the mlU host. The
1984
6617
ODMR frequencies and zfs assignments are given in Table 11. Additional signals are observed in the range below 3 GHz when the phosphorescence of the anticodon-anticodon complex is monitored at 1.2 K. The signal frequencies are given in Table 11. The two higher frequency transitions were assigned previously by Hoover et al. (1974) to yW in yeast tRNAPhe. The v 1 transition could not be observed in the tRNA or in the excised yW. Formation of the anticodon complex with E. coli tRNAzGIUinduces the appearance of the v I signal as well as a change in the polarity of the v 3 signal from a decrease to an increase in phosphorescence intensity. These effects have been observed previously by Luk (1975). Discussion The modified base mnm5s2Uis found at a strategic location in the tRNA molecule, namely the wobble position of the anticodon triad. Modification of this base is believed to have important biological implications, to the extent that the absence of such modifications is considered to be lethal (McCloskey 8i Nishimura, 1977). Minor uridine bases at the wobble position are believed to exercise control in codon recognition processes (Ishikura et al., 1971; Sekyia et al., 1969). An anticodon group containing s2U shows increased specificity of codon recognition in that base pairing with A is enhanced while that with G is excluded. This behavior is attributed to the reduced ability to sulfur to form hydrogen bonds (Yoshida et al., 1970; Agris et al., 1973). In addition, the side chain at position 5 is believed to have a major role in determining the three-dimensional structure of the anticodon loop (Berman et al., 1978; Sen & Ghosh, 1976; Hillen et al., 1978). Crystal structures of bases that have attached groups at the 5-position (Berman et al., 1978; Hillen et al., 1978) show that these have a strong influence on the hydrogen bonding of the 4-keto group of uracil. In addition, this group has been shown to alter the stacking pattern of neighboring groups on the base at the wobble position. For these reasons, the effect of binding the anticodon of E. coli tRNAZGluto the complementary anticodon of yeast tRNAPheon the ODMR properties of mnm5s2U at the wobble position is of great interest. The extremely large variation observed in the zfs of m1s2Uin various environments (Table I) indicates that it is an extremely sensitive probe of local environmental perturbations. The principal effects that we observe from the ODMR of mnm5s2Uin E . coli is (A) the existence of two major sites with differing zfs, each characterized by a large value of ID1 in the uncomplexed tRNA, and (B) the conversion of these sites to one having a smaller value of ID1 (comparable to that observed in m1s2U in the mlU host) when complexing with yeast tRNAPhetakes place. The zfs observed for uncomplexed E. coli tRNA,G1U is, in fact, comparable to those found for sites in neat m1s2U and for m1s2Uin ethyl acetate. We have shown previously (Taherian & Maki, 1981a; Taherian et al., 1982) that the large values of 1 0 1 that occur in thiouracils are a consequence of spin-orbit coupling induced by the presence of the sulfur atom. The increase in ID1 monotonically in the order U < s2U < s4U C s2s4U,which accompanies the increase in triplet-state decay constant in the same order, indicates that the heavy atom effect from sulfur at the 4-position is larger than that at the 2-position. In order for the heavy atom effect to be produced, the sulfur atom must participate in the delocalized r system that is characteristic of th excited T-T* state. We expect a more pronounced heavy atom effect from sulfur at position 4, since it is known (Hug & Tinoco, 1979; Nagata et al., 1973) that the lowest excited state of uracils is delocalized over the network O=C(4)-C-
6618
B I OCH EM ISTR Y
(5)=C(6). A recent investigation (Nishimura & Tsuboi, 1980) of resonance Raman effects in uracils and thiouracils has confirmed this description. In uracil, for instance, they found that the C(4)=0 vibration undergoes a resonance enhancement by the 260-nm absorption band, while the C(2)=0 vibration does not undergo resonance enhancement. The explanation is that the electronic excitation does not include the C(2)=0 group and thus that this chromophore is not conjugated with the one that is responsible for the electronic excitation. In the same work, Nishimura & Tsuboi (1980) show that the C(4)=S group of s4U is conjugated while the C(2)=0 group is not conjugated with the chromophore responsible for the absorption band at 333 nm. On the other hand, they show that the C(4)=0 group of s2U is conjugated with the chromophore absorbing at 300 nm, while the C(2)=S stretching Raman line (7 17 cm-’) appears weakly on excitation at 300 nm. According to these authors, “the pi-electrons on C(2)=S must migrate into the O=c(4)-C(S)=C(6) system to some extent, in contrast to the uracil C(2)=0, whose pielectrons do not seem to migrate into the O=C(4)--C(5)= C(6) system at all”. These observations are in close accord with our findings from the ODMR measurements on the thiouracil systems. The increase in the value of /Dl between U and m1s2Uin mlU (Table I) is consistent with some involvement of ?r electrons of C(2)=S in the delocalized triplet-state wave function. On the other hand, the further increase of ID1 when m1s2Uis placed in the ethyl acetate host or when it is studied as a neat solid is probably best explained in terms of the increased involvement of C(2)=S in the triplet wave function. Such an increased involvement also is characteristic of the mnm5s2U site in E . coli tRNAZGIU.It is possible that the mnm5s2U finds itself in a relatively disordered and solvent-exposed site similar to mlsZUin ethyl acetate. Formation of the anticodon-anticodon complex, on the other hand, produces a new environment in which the C(2)=S group is further isolated from the triplet-state chromophore. Judging from the zfs, this environment may be more homogeneous and similar to that of m1s2Uin the m’U host crystal. Apparently, the zfs of s2U are far more sensitive to subtle environmental perturbations than are those of s4U. Conformational changes in E . coli tRNAIVa’induce changes in the zfs parameters of s4U that are only a few percent (Taherian & Maki, 1981b). This is in contrast to the reduction of 1 0 1 by 15-25% that occurs in mnm5szUof E. coli tRNAZGluupon formation of the anticodon-anticodon complex. The reason for the enhanced sensitivity of s2U to environmental perturbations appears to stem from the partial involvement of the C=S chromophore in the triplet excited state. Small changes in this involvement induced by environmental effects can produce a relatively large effect on the zfs through the agency of spin-orbit coupling. In s4U, on the other hand, since the C=S group already is fully involved in the triplet excited state, environmental effects produce relatively small changes in the zfs. Registry No. m’s2U, 615-78-1; mnmss2U, 21263-85-4.
References Agris, P. F., Soell, D., & Seno, T. (1 973) Biochemistry 12, 4331-4337. Berman, H. M., Marcu, D., Narayanan, P., Fissekis, J. D., & Lipnick, R. L. (1978) Nucleic Acids Res. 5 , 893-903.
TAHERIAN, LUK, A N D MAKI
Clarke, R. H., Ed. (1982) Triplet State ODMR Spectroscopy: Techniques and Applications to Biophysical Systems, Wiley, New York. Crothers, D. M., Cole, P. E., Hilbers, C. W., & Shulman, R. G. (1974) J . Mol. Biol. 87, 63-88. Davis, J. M., & Maki, A. H. (1982) Proc. Natl. Acad. Sci. U.S.A. 79, 43 13-43 16. Eisinger, J. (1971a) Biochem. Biophys. Res. Commun. 43, 854-86 1. Eisinger, J. (1971b) Biochem. Biophys. Res. Commun. 44, 1135-1 142. ElSayed, M. A,, Owens, D. V., & Tinti, D. S . (1970) Chem. Phys. Lett. 3, 395-399. HtlBne, C., & Yaniv, M. (1970) Eur. J . Biochem. 15, 500-504. HElBne, C., Yaniv, M., & Elder, J. W. (1968) Biochem. Biophys. Res. Commun. 31, 660-664. Hillen, W., Egert, E., Lindner, H. J., & Gassen, H. G. (1978) Biochemistry 17, 5314-5320. Hoover, R. J., Luk, K. F. S., & Maki, A. H. (1974) J . Mol. Biol. 89, 363-378. Hug, W., & Tinoco, I., Jr. (1979) J . Am. Chem. SOC.95, 2803-28 13. Ishikura, H., Yamada, Y., & Nishimura, S . (1971) Biochim. Biophys. Acta 228, 47 1-48 1. Luk, K. F. S. (1975) Thesis, University of California, Riverside. Maki, A. H., & Co, T. (1976) Biochemistry 15, 1229-1235. Maki, A. H., Svejda, P., & Huber, J. R. (1978) Chem. Phys. 32, 369-380. McCloskey, J. A., & Nishimura, S. (1977) Acc. Chem. Res. 10, 403-410. Nagata, C., Imamura, A,, & Fujita, H. (1973) Adu. Biophys. 4, 1-69. Nishimura, S. (1972) Prog. Nucleic Acid Res. Mol. Biol. 12, 49-8 5. Nishimura, Y., & Tsuboi, M. (1980) Science (Washington, D.C.) 210, 1358-1360. Olmstead, J., & ElSayed, M. A. (1974) in Creation and Detection of the Excited State (Ware, W. R., Ed.) Vol. 2, pp 2-63, Marcel Dekker, New York. Pownall, H. J., Schaffer, A. M., Becker, R. S., & Mantulin, W. W. (1978) Photochem. Photobiol. 27, 625-628. Sekyia, T., Takeishi, K., & Ukita, T. (1969) Biochim. Biophys. Acta 182, 41 1-426. Sen, G. C., & Ghosh, H. P. (1976) Nucleic Acids Res. 3, 523-535. Singhal, R. P., & Fallis, P. A. M. (1979) Prog. Nucleic Acid Res. Mol. Biol. 23, 227-290. Taherian, M. R. (198 1) Thesis, University of California, Davis. Taherian, M. R., & Maki, A. H. (1981a) Chem. Phys. 55, 85-96. Taherian, M. R., & Maki, A. H. (1 98 1b) Biochemistry 20, 7295-7301. Taherian, M. R., Fink, W. H., & Maki, A. H. (1982) J . Phys. Chem. 86, 2586-2591. Vorbrueggen, H., & Strehlke, P. (1973) Chem. Ber. 106, 3039-3061. Yoshida, M., Takeishi, K., & Ukita, T. (1970) Biochem. Biophys. Res. Commun. 39, 852-857.
